Effects of Device Packaging and PC Board Materials on Radiation Dose in the Die

1. Effects of Device Packaging and PC Board Materials on Radiation Dose in the Die Edward R. Long, Jr. Longhill Technologies, inc. Waynesboro, VA 540-363-0104 LonghillTechnologies@PolyRAD.net 2004 MAPLD Ronald Reagon Building & International Trade Center Washington, DC September 8-10, 2004 Session P, Paper 150 P150

2. Presentation Outline • Aspects of spacecraft and on-board electronics that contribute to radiation shielding • What was modeled for total dose • A single electronics board populated with a 3 X 5 array of microelectronic devices • whose packaging are plastic or ceramic (with and without lids), • each of which are represented with several degrees of supplemental shielding • and the board is studied with and without a 4-oz copper ground plane • Dose-at-the-die for the array of devices for two board locations in two spacecraft designs • Three orbital environments: GSO, GTO, and SSO • Discussion of the results • Effects of packaging, PCB, and shielding on TID • Implications for dosimeter measurements P150

3. Aspects of Spacecraft and On-board Electronics That Contribute to Radiation Shielding • Aspects discussed in an earlier presentation1 • Board location • Where on the spacecraft • Which board in an ordered stack of boards • Board Orientation • Device position • Location on the board • Adjacent devices • The structure, materials, and location of other on-board components • Aspects discussed in this presentation • Device packaging materials • Plastic • Ceramic • PCB materials • No ground plane • 4-0z Cu ground plane • Supplementary shielding • 10-mil, 20-mil, 40-mil, and 80-mil PolyRAD • Two inward-facing board locations (between –X shear panel and –X battery) • Near –X shear plane • Near –X battery 1. - Long, Edward R., Jr., Modeling the Requirements for Radiation Shielding, 2003 MAPLD International Conference, Washington, DC, September 9-11, 2003 P150

4. Exterior instruments Solar panel Instrument support panel -X Battery Anti-buckle panel Tank support panel Tanks Internal component Shear panel Description of the Dose Modeling • Two spacecraft designs • A Generic spacecraft1 • Three-Quarter view • Intrinsic shielding equivalent to 0.135-in aluminum • Thin-walled version of the generic spacecraft • Intrinsic shielding equivalent to 0.040-in aluminum 1. - Long, Edward R., Jr., Modeling the Requirements for Radiation Shielding, 2003 MAPLD International Conference, Washington, DC, September 9-11, 2003 P150

5. Column 2, Column 2, Column 3, Column 3, (alumina (alumina (alumina (alumina Column 1 Column 1 0 0 - - oz or 4 oz or 4 - - oz oz encapsulate, encapsulate, encapsulate, encapsulate, (plastic (plastic copper copper with with kovar kovar without without encapsulate) encapsulate) ground ground lid) lid) kovar kovar lid) lid) plane plane kovar lid kovar lid die die 0.062 0.062 - - in FR4 in FR4 Non - shielded Non - shielded Row Row 10 10 - - mil mil PolyRAD PolyRAD q q Shielded row Shielded row 20 10 - - mil mil PolyRAD PolyRAD q q Shielded row Shielded row 40 10 - - mil mil PolyRAD PolyRAD q q Shielded row Shielded row 10 80 - - mil mil PolyRAD PolyRAD q q Shielded row Shielded row Sketch is not to scale Description of the Dose Modeling (continued) • Orbits • GSO: 35800 km x 35800 km x 0 deg • GTO: 36000 km x 300 km x 7 deg • SSO: 800 km x 800 km x 98 deg • Single electronics board consisting of a 3 X 5 array of copies of a device, modeled with plastic and ceramic encapsulation (packaging). P150

6. 0.705” 0.197” 0.208” 0.300” 0.157” 0.042” Y 0.201” 0.123” 0.295” Kovar lid: 0.455” x 0.274” x 0.010” X 0.039” x 0.039” 0.027” 0.087” Z 0.027” X Description of the Dose Modeling (continued) • Cross section of modeled device, shown using ceramic encapsulation When in place, the lid is centered on the top of the device Top and side views (Kovar lid removed) P150

7. Discussion of the Results • External and internal annual doses for solar maximum conditions1 in three orbit typesand a PCB with no (0-oz) ground plane • Comments • The additional shielding provided by the 135-mil spacecraft varies with the energy distributions of the electrons and protons within the orbit. It is a factor of approximately 30 for GSO, 15 for GTO, and 8 for SSO. • The effect of the location in the spacecraft (at the negative-X axis’s battery or shear panel) on the dose is a complex consideration that depends on the exposure’s solid angle, exterior panel thickness, and adjacent interior components. • The PCB (without a 4-Oz ground plane on its backside) reduces the dose by an additional factor of 2. Although this amount of dose reduction is not sufficient in most cases it does indicate that the board material and its sizing should be included as part of the dose modeling. 1.: Trends for solar minimum conditions are the same 2 : 40-GSO-0Oz-Batt represents the 40-mil spacecraft in GSO having an electronics board that has no rear ground plane and located near the –X battery 3: An external location on the top of the spacecraft 4: An internal location inside the Spacecraft next to the +Z,-Y corner of the electronics board 5. A location on the populated face of the electronics board, near the +Z,-Y corner of the electronics board NOTE: The board’s populated side faces the +X direction, i.e. towards the center of the spacecraft. P150

8. Discussion of the Results (continued) • Annual dose on the surface of the board, with and without a ground plane (1), for the two board locations (2), and the three orbits. • – “0oz” means without ground plane and “4oz” means 4-oz ground plane • - “Batt” means located near the battery and “Shear” means near the shear panel • Comments • The dose reduction due to board location (outboard near the battery versus inboard near the shear panel) is larger for the 40-mil spacecraft and varies more with the orbit. Thus PCB’s contribution to the total shielding is more significant for smaller spacecraft. • The additional shielding provided by the 4-Oz copper ground plane is not consequential for either the 40-mil or the 135-mil spacecraft. P150

9. Discussion of the Results (continued) • Dose reduction effects of plastic and ceramic encapsulation 40-mil Spacecraft 135-mil Spacecraft • Comments • Dose modeling typically models the dose inside the spacecraft (bar 1) and doesn’t include device encapsulation • For the 40-mil spacecraft, plastic encapsulation (bar 2) reduces the dose an additional order of magnitude in all 3 orbit types. The ceramic encapsulation (bar 3) reduces the dose 2 additional orders of magnitude, except for the SSO. • For the 135-mil spacecraft, the plastic encapsulation (bar 2) reduces the dose by a factor of ~5 and the ceramic encapsulation (bar 3) reduce the dose an additional order of magnitude. • The trends for the effects of encapsulation are also true for near the shear panel and for solar minimum conditions at both locations. (data not shown) • The additional shielding provided by encapsulation may explain flight cases for which parts do not fail even though gamma radiation tests suggest the opposite. P150

10. Discussion of the Results (continued) • Additional dose-reduction provided by PolyRADq1 for plastic encapsulated (packaged) die 1 – PolyRAD is an advanced radiation shielding material developed under a NASA SBIR contract • Comments • These data are for plastic encapsulated (packaged) die adjacent to the –X battery inside the 40-mil spacecraft. Conventional representation of the dose reduction provided by a supplemental shielding, as shown in the next slide, does not include the encapsulation material. • The limit for dose reduction is achieved by a shield thickness of approximately 20 mils. This is 25% of the thickness indicated by conventional representations (see the next slide) and thus a significant weight savings. P150

11. Discussion of the Results (continued) • Conventional methods for representations of the dose-reduction provided by PolyRADq1 No spacecraft 40-mil spacecraft 1 – PolyRAD is an advanced radiation shielding material developed under a NASA SBIR contract • Comments • On the left: Just PolyRAD, thus the dose for zero thickness is that of bare (neat) space • On the right: PolyRAD + 40-mil aluminum, thus the dose for zero thickness is that behind 40-mil aluminum. • The limit for dose reduction requires from 80- to 160-mil thick PolyRAD. Thus conventional modeling may overestimate the amount of required supplemental shielding. This could be a error for small, weight-critical spacecraft. (Conventional modeling does not account for shielding provided by the spacecraft/internal component structure and the material type and size for electronic boxes, PCB, and die encapsulation.) P150

12. Discussion of the Results (continued) • Dosimetry implications of encapsulation • Comments • The ceramic-packaged die in this study is similar to a candidate radfet package for space flight dosimetry. • The measured dose predicted by the modeling (kovar lid in place) is from 1 to 2 orders of magnitude less than on the board surface: • For 40-mil spacecraft – Modeled measured dose is from 1 to 8 percent of that at the board • For 135-mil spacecraft – Modeled measured dose is from 1 to 4 percent of that on the board • Removal of the Kovar lid significantly reduces the dosimetry error. P150

13. Summary Remarks • This study modeled the dose-at-the-die for an array of copied devices on a single board at two locations in two different generic spacecraft designs (40-mil and 135-mil) for GSO, GTO, and SSO radiation environments. • Observations from the modeled doses: • PCB (0.062”) reduced the dose approximately 50 % • 4-oz Copper ground planes did not have a significant effect on the dose. • For the 40-mil design the location of the board varied the dose as much as 50%. • The combined effect of the spacecraft structure, internal adjacent components, the PCB and its location, and the device encapsulation reduced the dose two orders-of magnitude for GSO and GTO and a little more than one order-of-magnitude for the SSO. • The amount of supplemental shielding (in this study PolyRAD) to attain low doses using the detailed approach to modeling in this study is 25 % or less than that determined from conventional modeling. This indicates that significant savings in the mass cost of supplemental shielding are possible. • The effects of encapsulation material can introduce significant dosimetry errors for devices such as radfets. P150